High availability cluster server for enterprise data management

Abstract

A high availability, scalable cluster server system for enterprise data management. The server is a cluster of two or more nodes. Each node runs one or more virtual servers. A virtual server consists of network resources and resources for enterprise data management. Failover is based on moving virtual servers from a failed node to a healthy node. The invention provides for network failover within the same node. Enterprise data management consists of data discovery, data categorization and applying enterprise policies on categorized data. One of the problems of data management is the diversity of policies and in some cases their conflicting requirements. According to one aspect of the invention, enterprise policies are broking into policy entities. Policy entities represent the simplest policy unit. Some of the policy entities are shared between the diversified enterprise policies. Identifying a data management policy with the highest priority and applying the policy entity that belongs to it resolve conflict in policy requirements.

Description

BACKGROUND OF THE INVENTION[0001]The invention relates to a high availability and scalable cluster server for enterprise data management.[0002]High availability cluster server is a server that continues to function even after a failure of system hardware or software. The usual way of providing high availability is to duplicate system components. If some component becomes unavailable, another can be used instead. Scalable cluster server is a server that is able to increase performance and workload by adding more hardware or software resources.[0003]A cluster is a group of servers and other resources that act like a single system and enable high availability and load balancing. The servers are referred to as nodes. Nodes typically consist of one or more instruction processors (generally referred to as CPUs), disks, memory, power supplies, motherboards, expansion slots, and interface boards. In a master-slave design, one node of the system cluster is called the primary or master server...

AI Technical Summary

Benefits of technology

[0029]Implementations of the invention can realize one or more of the following advantages. Failover used only as a last resort, and consequently the disruption caused by failover to the accessibility of services is limited. Total system performance is improved through load balancing.

[0030]In general, in another aspect, the invention provides systems, programs, and methods that allow replicating metadata and search index repositories belonging to one node into another node or nodes. This provides redundant repositories. In case the node that owns the repository failed, its virtual servers are migrated to where the repositories are replicated. The master makes sure that the workload of the nodes that contain the replicated repositories is not high to ensure that they can handled extra work when virtual servers are migrated to them during failover. This alleviates the problem of re-creating the metadata and the search index of the failed node after failover.

[0032]In general, in another aspect, the invention provides systems, programs, and methods for expanding the storage in nodes by adding more drives. This enables the system to allocate more storage for the search index and the metadata repositories and the local database where the results of data management are stored. As a consequence the cluster can manage larger amount of enterprise data.

Problems solved by technology

One of the challenges for data management products is to provide solutions to different aspects of data management in one platform.

One of the challenges to data management is the exponential growth of the enterprise data.

Another challenge is diversity of devices where data exists.

Another problem domain is the proliferation of data into the fringes of the enterprise network, namely laptops and remote users.

Images